The purpose of this project is to study epigenetic alterations in CML. We have previously discovered imprinted genes in humans and loss of imprinting (LOI) in cancer. Genomic imprinting is a chromosomal modification in the gamete or zygote leading to differential expression of the two parental alleles of a gene in somatic cells of the offspring. LOI can lead to expression of the normally silent copy of growth-promoting genes in cancer, and los of expression of the normally transcribed copy of growth inhibitory genes. In addition, we have found alterations in DNA methylation, a covalent modification of the nucleotide cytosine, that are specific for tumor cells with LOI. Collaborating with this program, we have discovered LOI in CML, which appears to be specific for progression to accelerated phase and blast crisis. This is important because, while a great del is known about the causes of stable phase, little is known about progression to blast crisis, the cause of significant morbidity and mortality in CML. We will extend our preliminary observation to a large number of patients at varying stages of disease progression, and determine its frequency and specificity for disease stage. We will determine which genes show LOI in CML, the effect of LOI on their expression, and the role of DNA methylation in this process. Based on other studies from our laboratory showing that normal imprinting can be restored to some tumor cells with LOI using 5-aza-2'-deoxycytidine, we will determine whether 5- aza-2'-deoxycytidine treatment in vitro or in vivo leads to restoration of normal imprinting in CML cells with LOI. We will also determine the relationship between LOI and DNA methylation of the affected genes and other loci. This project will also address the molecular mechanism of altered DNA methylation in CML progression. DNA methylation affects genomic imprinting, and abnormal methylation of genes such as bcr and abi is a feature of CML progression. We originally discovered altered DNA methylation in cancer, and we have found that the only identified genes that regulates DNA methylation, cytosine DNA methyltransferase, is transcribed at normal levels in tumor cells, suggesting that its activity may be modulated by other factors. Consistent with this hypothesis, we have identified a gene whose protein product interacts with DNA methyltransferase and may modulate its activity in cells. We will characterize these genes, determine the effect of its product on DNA methylation in transfection experiments, and look for alterations in its structure or regulation in CML. These experiments may provide direct insight into the pathogenesis of CML progression and/or novel therapeutic targets or reagents for the therapy of CML.